Don't have metrics about GPU Utilization

[tsvyake]

Hello Dear all, I encountered a problem specifically the parameter "GPU utilization" does not show, I tried three templates, all as one show the temperature aand how much GPU memory is used and free. All these three options of templates and scripts, including the one below, work perfectly on Ubuntu and show all the parameters




I have Zabbix 7 Version
System from which I take readings
Debian 12 (6.1.0-28-amd64)
Zabfix agent2

config (for include)

UserParameter=gpu.number,/usr/bin/nvidia-smi -L | /usr/bin/wc -l
UserParameter=gpu.discovery,/etc/zabbix/scripts/get_gpus_info.sh
UserParameter=gpu.fanspeed[*],nvidia-smi --query-gpu=fan.speed --format=csv,noheader,nounits -i $1 | tr -d "\n"
UserParameter=gpu.power[*],nvidia-smi --query-gpu=power.draw --format=csv,noheader,nounits -i $1 | tr -d "\n"
UserParameter=gpu.temp[*],nvidia-smi --query-gpu=temperature.gpu --format=csv,noheader,nounits -i $1 | tr -d "\n"
UserParameter=gpu.utilization[*],nvidia-smi --query-gpu=utilization.gpu --format=csv,noheader,nounits -i $1 | tr -d "\n"
UserParameter=gpu.memfree[*],nvidia-smi --query-gpu=memory.free --format=csv,noheader,nounits -i $1 | tr -d "\n"
UserParameter=gpu.memused[*],nvidia-smi --query-gpu=memory.used --format=csv,noheader,nounits -i $1 | tr -d "\n"
UserParameter=gpu.memtotal[*],nvidia-smi --query-gpu=memory.total --format=csv,noheader,nounits -i $1 | tr -d "\n"
UserParameter=gpu.utilization.dec.min[*],nvidia-smi -q -d UTILIZATION -i $1 | grep -A 5  DEC | grep Min | tr -s ' ' | cut -d ' ' -f 4
UserParameter=gpu.utilization.dec.max[*],nvidia-smi -q -d UTILIZATION -i $1 | grep -A 5  DEC | grep Max | tr -s ' ' | cut -d ' ' -f 4
UserParameter=gpu.utilization.enc.min[*],nvidia-smi -q -d UTILIZATION -i $1 | grep -A 5  ENC | grep Min | tr -s ' ' | cut -d ' ' -f 4
UserParameter=gpu.utilization.enc.max[*],nvidia-smi -q -d UTILIZATION -i $1 | grep -A 5  ENC | grep Max | tr -s ' ' | cut -d ' ' -f 4

script get_gpus_info.sh

#!/bin/bash

result=$(/usr/bin/nvidia-smi -L)
first=1

echo "{"
echo ""data":["

while IFS= read -r line
do
  if (( "$first" != "1" ))
  then
    echo ,
  fi
  index=$(echo -n $line | cut -d ":" -f 1 | cut -d " " -f 2)
  gpuuuid=$(echo -n $line | cut -d ":" -f 3 | tr -d ")" | tr -d " ")
  echo -n {""{#GPUINDEX}"":"$index"", "{#GPUUUID}"":"$gpuuuid"}
  if (( "$first" == "1" ))
  then
#    echo ,
    first=0
  fi
done < <(printf '%s\n' "$result")

echo
echo "]"
echo "}"

[Jan_Prusinowski]

Have you made sure that the GPU driver is installed?
Please try running the following command:

sudo apt install nvidia-driver nvidia-smi

If this doesn't fix the issue, please provide logs that might indicate the problem.

[tsvyake]

Hello ,yes drivers all has been installed ,we can use GPUs

pls see

nvidia-smi
Thu Dec 12 09:31:58 2024
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 560.35.03 Driver Version: 565.57.01 CUDA Version: 12.7 |
|-----------------------------------------+------------------------+----------------------+